The OYEE lab / Free / Hands on
The best way to understand our approach? Try it. Explore three mechanisms at your own pace. In a workshop, an educator guides these discoveries with your group.
Teaching simulations: no live AI. Progress stays on this device; no entered text is saved or sent.
Experiment 01 / Prediction isn’t knowledge
A language model assigns probabilities to upcoming tokens. Here, three words and illustrative weights make that mechanism visible.
Temperature reshapes the probabilities. It measures neither truth nor intelligence.
Which word will emerge?
A likely continuation is not evidence. The same machine can produce a fluent sentence and a convincing mistake. Try two contexts and two temperatures, then sample a word.
Experiment 02 / One answer, three checks
Read the fictional source document. Then classify each claim in the summary. Missing information is not automatically false.
The team proposes a 90-minute workshop to explore AI. The room can hold 24 people. The date is yet to be confirmed. Free registration is being considered, but the budget has not yet been approved.
Your only source for this experiment.
Return to the source. Distinguish “supported”, “contradicted” and “not enough information”. This is the reasoning we practise together in a workshop.
Experiment 03 / Make your intent visible
Turn “tell me about AI” into a request someone can actually follow. Build, compare and copy your instruction.
Use a fictional example. This text stays on the page.
“Tell me about AI.”
Your instruction, assembled as you goThis assembles your instruction; no AI response is generated.
A precise instruction makes the result easier to evaluate. It does not guarantee accuracy. Choose your audience, task and at least two criteria to finish this experiment.
You explored prediction, checked claims and built an instruction. Imagine the conversations this can spark with your group.
Bring this to your workshop Continue the free journey ↗A little look inside / 002
Let’s open the box. Explore the journey of an answer, from the first word to your decision.
Click a layer or choose a stage.
Explain AI to my team.
It starts with you. A question, a context, an intention.
Explain · AI · to · my · team
Text becomes tokens: parts of words, words or symbols. The boundaries shown here are simplified.
tool / system / technology
The model calculates possible continuations from context. A likely continuation can still be wrong.
Is it accurate? Useful? Suitable?
You compare against sources, identify limitations and decide what to use. Your judgment remains essential.
Teaching view: these layers represent stages, not a model’s physical architecture. Preset examples; no live AI.
The next experiment is with you.
Questions fly. Ideas meet. An educator guides the room. That’s an OYEE workshop.
Let’s plan your workshop